Check Fixture Validation

[plopperzz]

Good morning everyone! I have a fixture here that I need to check but its set up is a little different than i am used to as the CAD is not level to any of the planes in the coordinate system, yet the part on the fixture is level to the late holding the tooling balls.

I find this to be an issue because if I transform the part coordinates to level out the CAD, then the tooling ball coordinates on the fixture won't be correct, and it i try to change the STARTUP alignment I seem to mess the tooling balls up as well. Anyone have a solution for this?

The CAD, as imported, is in the general correct orientation as it is in car position, but on an approximate 5° angle from the x-axis to the -z-axis so when programming the part that I had to do was a slight rotational transform to program/measure the part. The part as it sits on the fixture, however, is upside down and level to the ground. I do have the CAD for the fixture, but I am under the impression that I must validate the fixture using the part CAD and not the fixture CAD.

Any help is much appreciated.

[RandomJerk]

You should read up on offset planes - this will allow you to measure the "skewed" tooling balls but level them square.

Fundamentally, when you create the plane you select the balls, but then you tell them what the height differences are supposed to be. It will then run the plane with those offsets.

[VinniUSMC]

Iterative alignment to the tooling balls. Then iterative alignment to the fixture datums. Plenty of threads here about doing that.

No need to do any transforming, rotating, whatever. It's supposed to stay in car body anyway.

[VinniUSMC]

With iterative alignments, the orientation of the CAD is irrelevant. How you align it on the machine is how PC-DMIS will see it aligned, no matter what orientation the part is in in car body.

[plopperzz]

VinniUSMC When I try this the tooling ball coordinates get shifted away from the part, even though there is no alignment (other than STARTUP) governing their coordinates. The same thing happens if I try to do a regular alignment leveling to a plane constructed from the tooling balls...

[VinniUSMC]

I don't know what you're doing, but iterative alignments don't change relative locations, so what you're doing is something else. Post your code.

[plopperzz]

You should read up on offset planes - this will allow you to measure the "skewed" tooling balls but level them square.

Fundamentally, when you create the plane you select the balls, but then you tell them what the height differences are supposed to be. It will then run the plane with those offsets.

I have tried that, and I'm not sure what I'm doing wrong. but here is what I'm dealing with.

This is how the cad loads.
1.pngHere is how the part sits in the fixture.2.pngThis is what happens any time I run an alignment...3.png

[RandomJerk]

@plopperpzz - how are you programming the spheres? Are you selecting them from the model, or are you touching off on the machine? It looks like you're touching off the spheres and not updating the THEO values to match the CAD THEO values.

If you were to bring in the model and select a sphere in CAD, it will spit out the THEO coordinates relative to the CAD zero. Pretend they're <1,1,1> to make it easy.

If you grab the RCU and touch off that sphere, you'll get THEO coordinates relative to wherever the current ACTUAL alignment is, they'll probably be something like <10.4453,-15.6843,-11.3809>. What happens is when you then zero to that sphere, you're telling the CAD model to set the zero to those theoretical coordinates, but since it only knows it's (CAD) zero, it will do so in CAD coordinates. This throws everything else off. If you're going to pick things directly from the part on the machine, you need to adjust the THEOs to the correct (CAD) nominals before doing anything with the alignment.

[plopperzz]

RandomJerk I am putting in the tooling balls theoretical values in autospheres for their location, and surface vectors of <,0,0,1>, and when I run it, this happens.

[Matthew D. Hoedeman]

NEW PROGRAM
IMPORT CAD DATA
CTRL-F3 (rotate cad VIEW, not the cad itself) until it looks as it sits on the machine as you look down from above)
F5 ~ PART/MACHINE tab: set the machine axis to match the cad axis (THIS DOES NOT CHANGE THE CAD DATA IN ANY WAY!)
CTRL-F (autofeatures) select SPHERE. Type in the XYZ IJK values for the sphere
Create the sphere.
Repeat for the other 2
construct an ITERATIVE alignment selecting 3 spheres, then 2 spheres, then 1 sphere and done
Execute the program